“According to two common measures of map fairness, congressional maps enacted by commissions (or courts that took over from failed commissions) have been less biased than those that have emerged from legislatures. For instance, out of the six commission states with at least three congressional districts, five have a median seat whose FiveThirtyEight partisan lean2 is within 3 percentage points of the state’s as a whole. (The exception is Colorado, where the median seat is 5 points redder than the state.)
It’s even more striking when you go by the maps’ efficiency gaps, which is a measure of which party has fewer “wasted” votes (i.e., votes that don’t contribute toward a candidate winning). All but one commission state with at least three congressional districts has an efficiency gap of 5 points or fewer, whereas the maps drawn by partisan actors are very partisan. (So far, every Democratic-controlled state with at least three districts has an efficiency gap of D+13 or greater, while all but one Republican-controlled state with at least three districts has an efficiency gap of R+7 or greater.)
The exception among commission states is New Jersey, whose map has a D+16 efficiency gap, indicating a strong pro-Democratic bias. But New Jersey’s commission is not exactly a model of nonpartisanship. Twelve of its 13 members are picked directly by state legislators or political parties (six by Democrats, six by Republicans), and after they failed to agree on a 13th member last summer, the New Jersey Supreme Court chose the Democrats’ preferred candidate. The commission eventually (and predictably) voted 7-6 for a map drawn by the commission’s Democrats.”